The Generic Characteristic data type is used for predictive variables that will not be grouped. The number of bins in the analysis will be the number of unique values for this characteristic within the data.
Qualitative data, and more specifically columns made up of text values, typically fall under the Generic Characteristic data type, since ordering the data alphabetically does not usually result in values whose neighboring values are more related than the values farther away. This is not to say that numerical data cannot be a generic characteristic; for example, a "Group Number" column would make more sense as a generic characteristic if group 2 is no more related to group 3 than it is to group 99. Also, a numeric column with a reasonably small number of different values may be more valuable as a generic characteristic than as a grouped characteristic.
When Generic Characteristic is chosen as the data type and stronger relationships are expected between neighboring values than between values farther away, Smoothing the data across bins may be desirable.